Lower Lakes Aquatic Vegetation Survey Project

Transcription

1 Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project Prepared for the Basin Environmental Improvement Project Commission By the Coeur d Alene Tribe Lake Management Department December 2007

2 Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project Prepared for the Basin Environmental Improvement Project Commission By the Coeur d Alene Tribe Lake Management Department 401 Annie Antelope Road Plummer, Idaho December 2007

3 Table of Contents Executive Summary... 1 Acknowledgements... 4 Introduction... 5 Description of Study Area... 6 Purpose / Objectives... 8 Study Purpose... 8 Study Objectives... 8 Materials and Methods... 9 Transect Sample Collection Techniques... 9 Transect Sample Sorting Techniques Grid Sampling Methods Laboratory Analyses Biomass Analyses Total Phosphorus Content Total Nitrogen Content Nutrient Loading Calculations Quality Assurance and Quality Control Results General Plant Community Structure Grid Point Results Transect Results Biomass Results Nutrient Results QC Results Duplicates Nutrient Matrix Spikes Nutrient Reference Samples Nutrient Blanks Discussion Grid Results Versus Transect Results Comparison of Project Results with Those of Other Studies Biomass Nutrient Content Literature Review of Potential Nutrient Release from Macrophytes Overview Phosphorus Release Nitrogen Release Estimate of Nutrient Loading from Aquatic Vegetation Overview Aquatic Plant Growth Regions Calculation of Nutrient Pool Available for Potential Release Nutrient Release Criteria Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project i

4 Lower Lakes Nutrient Loading Result Combined Lower Lakes and Coeur d Alene Lake Nutrient Loading Summary Conclusions and Recommendations References Appendix A. Field and Laboratory Data from Transects... 1 Appendix B. Field Data from Grid Points... 1 Appendix C. Quality Control Results... 2 Appendix D. Project Photographs... 3 Appendix E. Nutrient Loading Calculation Spreadsheet... 4 List of Figures Figure 1. Coeur d'alene Lake, Idaho, showing Lower Lakes Aquatic Vegetation Study Area Figure 2. Map of Coeur d Alene Tribe s Lower Lakes Aquatic Vegetation Survey study area showing grid points and transect locations... 8 Figure 3. Quadrat and mesh bag used for aquatic vegetation sampling portion of Lower Lakes Aquatic Vegetation Survey project... 9 Figure 4. Weed rake used for grid sampling portion of Lower Lakes Aquatic Vegetation Survey project Figure 5. Location of submersed and emergent plant types, as well as areas where no plants were found, from the grid point sampling, Lower Lakes Aquatic Vegetation Survey Project Figure 6. Average aquatic vegetation biomass variations (all species) by transect for the Lower Lakes Aquatic Vegetation Survey project Figure 7. Average aquatic vegetation biomass variations (all species) by transect and year for the Lower Lakes Aquatic Vegetation Survey project Figure 8. Aquatic vegetation biomass variations by depth for all transects sampled for the Lower Lakes Aquatic Vegetation Survey project Figure 9. Location of aquatic plant growth regions used for nutrient release estimation for the Lower Lakes Aquatic Vegetation Survey project List of Tables Table 1. Data Quality Indicators applicable to the Coeur d Alene Lake Baseline Aquatic Vegetation Survey project Table 2. Description of transects sampled during the Lower Lakes Aquatic Vegetation Survey project Table 3. Species codes and scientific and common names of plants sampled during the Table 4. Summarized aquatic vegetation biomass statistics by species for transects sampled during the Lower Lakes Aquatic Vegetation Survey project Table 5. Summarized aquatic vegetation biomass by transect for the Lower Lakes Aquatic Vegetation Survey project Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project ii

5 Table 6. Summarized aquatic vegetation biomass by depth for all transects sampled during the Lower Lakes Aquatic Vegetation Survey project Table 7. Total phosphorus (TP) data collected for the Lower Lakes Aquatic Vegetation Survey project Table 8. Total Kjeldahl nitrogen (TKN) data collected for the 2005 Baseline Coeur d'alene Lake Aquatic Vegetation Survey project Table 9. Nitrate nitrogen (NO 3 ) data collected for the 2005 Baseline Coeur d'alene Lake Aquatic Vegetation Survey project Table 10. Aquatic plant growth regions established for the Lower Lakes Aquatic Vegetation Survey project Table 11. Estimated total phosphorus and total nitrogen loading from aquatic vegetation in the Lower Lakes project area based on 2004 and 2005 transect sampling and literature reported nutrient release rates Table 12. Predominant nutrient loading sources to Coeur d'alene Lake from 1991 and 1992 (from Woods and Beckwith, 1997) Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project iii

6 Executive Summary Nutrient management has been a key concern of the Coeur d Alene Tribe for many years. This is due to the increase in productivity and concurrent decrease in water quality that excessive levels of phosphorus and nitrogen can cause. Increased productivity can directly result in lower hypolimnetic oxygen levels which can, in turn, cause the release of toxic heavy metals which have accumulated in the sediments of Coeur d Alene Lake. The cleanup of metals contamination in soils and waters within the Coeur d Alene River and Lake system is one of the primary goals of the Basin Environmental Improvement Project Commission (BEIPC). A related issue in Coeur d'alene Lake area is the potential for infestation by the noxious, invasive plant Eurasian watermilfoil (Myriophyllum spicatum). This plant has been found in most of the nearby lakes in Idaho and Washington and is known to be introduced into un-infested waters via fragments which can be carried on boats and boat trailers. This plant represents a significant threat to lake beneficial uses including fish and wildlife habitats. The 2002 Clean Water Act appropriation for the US EPA included a line item for a grant to the State of Idaho (BEIPC) to undertake activities that conduct and promote the coordination and acceleration of research, investigations, experiments, training, demonstrations, surveys and studies relating to the causes, effects, extent, prevention, reduction and elimination of pollution. Thus, a Coeur d Alene Lake aquatic vegetation survey project, including an effort to estimate nutrient release from existing aquatic plant populations, was considered appropriate for funding through the BEIPC research and demonstration project. In particular, the southern end of Coeur d Alene Lake, an area referred to collectively as the Lower Lakes, and more specifically as Chatcolet, Round and Benewah Lakes, was chosen for the initial (pilot) phase of this survey because of the presence of extensive shallow water habitats with nutrient rich sediment substrates; i.e. prime aquatic plant growth areas. Following submission of a project proposal to the BEIPC Leadership Group review process and approval by the BEIPC Board, funding through the EPA Clean Water Act program was approved and a Sub-Grant Agreement was issued to the Coeur d Alene Tribe on June 1, The primary purpose of this survey was to develop baseline data on submersed aquatic plant species distribution and biomass in the three Lower Lakes. The secondary purpose was to estimate nutrient (phosphorus and nitrogen) release from the existing plant beds into the water column of these lakes and, subsequently into Coeur d'alene Lake. The tertiary purpose was to inspect these lakes for the presence of invasive, noxious aquatic species. Specific survey objectives were to perform SCUBA diver collection of submersed aquatic vegetation species

7 The results of the grid sampling at 194 and 197 points for the two years of this project provided an initial estimate of plant community structure. The average number of plant species seen per grid sample was 2.9 in 2004 and 2.5 in In both years the top three species groups found, in terms of the number of samples where they were present, were: Elodea (Waterweed) species, Myriophyllum (Milfoil) species and Potamogeton (thin leafed pondweed) species. The predominant Milfoil seen was the noxious aquatic weed The community structure seen in the 24 transects sampled, which were more efficient in showing all species present, was analyzed based on which of the three lake areas within the study area they were found in. In Benewah Lake the average number of species per sample was 3.3 in 2004 and 2.9 in The most predominant species in terms of frequency of occurrence in Benewah Lake were Elodea, Myriophyllum and Potamogeton robinsii (Fern-Leaf pondweed). In Chatcolet Lake the average number of species per sample was 3.0 in 2004 and 3.2 in The predominant species groups in Chatcolet Lake during both years were Elodea, Myriophyllum the thin leafed pondweeds. Round Lake exhibited the highest number of species per sample with an average of 3.9 in both 2004 and The three species groups Elodea, Myriophyllum and thinleafed pondweed were roughly equally predominant throughout this area with Ceratophyllum demersum (Coontail) and Potamogeton richardsonii (Richardson s pondweed) associated with these others. It appears that the plant biomass levels seen with this project are within the range of those seen elsewhere in North America and considerably less than those in euthrophic waters. The two frequently found plants with the highest mean biomass were Elodea species and the thin leafed Potamogeton species, which had very similar mean biomass levels of g/m 2 and g/m 2, respectively. The two species found at the highest individual sub-sample biomass were Coontail, at 1, g/m 2 and Elodea species at g/m 2. The biomass data compiled by transect (all species and depths) indicated that the highest average sample biomass was g/m 2 seen at Transect B5 (Benewah Lake). The lowest per sample biomass of g/m 2 was found on Transect C2 (Chatcolet Lake). Biomass summarized by depth (across all transects and species) indicates that submersed plants were found at depths between 1 and 18 feet in this study. There is an apparent progression in biomass between the 2 and 12 foot depths, with the highest sample biomass of g/m 2 at 12 feet, and then a rapid decline to the 18 foot depth. Nutrient analyses were performed on 13 species collected from the sampled transects and included analyses for total phosphorus (TP), total Kjeldahl nitrogen (TKN) and nitrate nitrogen (NO 3 ). Of the six predominant plant species that were found during this study, the highest mean TP concentrations were 4,084 and 4,055 μg P/g, with Elodea and Coontail, respectively. The lowest mean phosphorus concentration of the plants analyzed was 2,709 μg P/g in Potamogeton richardsonii. The organic forms of nitrogen (including ammonium), as measured by the Total Kjeldahl Nitrogen (TKN) analysis, were by far the predominant forms of nitrogen in all of the vegetation samples. Again, looking at the six primary species, the range of mean TKN values was only between 22,176 and 27,757 μg N/g and the standard deviation values were substantially below the mean values, an indication of lower variations in the individual analyses. The nitrate Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 2

8 (NO 3 ) concentration data, while only a small fraction of the TKN concentrations, were somewhat variable across the species tested. Effects of submersed aquatic macrophytes on nutrient cycling in lakes are most pronounced in shallow areas that support extensive stands of robust submersed species, such as the large species of Myriophyllum, Potamogeton and Elodea (Carpenter, 1980). These species have high biomass turnover during the growing season and therefore recycle nutrients when water temperatures are high and potential effects on plankton production are maximal (Carpenter, 1983). This being said, the estimation of nutrient release from a diverse plant assemblage in a lake the size of Coeur d Alene, or even the Lower Lakes study area, is a complex process and involves many assumptions to expand the limited available data. The estimated phosphorus and nitrogen loading from the Lower Lakes project area, based on the 2004 and 2005 transect sample analyses, and nutrient release criteria from the scientific literature, was calculated on the basis of aquatic plant growth region within the study area. The 219 acres of submersed plant habitat in Benewah Lake yielded 740 kg of phosphorus and 3,535 kg of nitrogen (both annual loading rates). In Chatcolet Lake, the three regions together which provided 848 acres of habitat were estimated to release 3,323 kg of phosphorus and 12,011 kg of nitrogen annually. Round Lake with 891 acres could release 3,778 kg of phosphorus and 12,522 kg of nitrogen. The resultant combined loading was 7,841 kg of phosphorus and 28,068 kg of nitrogen. On a per-acre basis, Round Lake had the highest phosphorus release at 4.2 kg per acre per year while Benewah Lake had the highest nitrogen rate at 16.1 kg per acre per year. The results of the Lower Lakes Aquatic Vegetation Survey project, which are similar in scope to those of the Coeur d Alene Lake project conducted for Avista Utilities, were combined so that a more complete picture of aquatic plant growth and potential impacts on lake water quality can be considered by the lake managers. The estimated, combined, annual nutrient loading from Coeur d Alene and the Lower Lakes resulting from these two similar studies is 13,444 kg phosphorus and 46,305 kg nitrogen. The loading of the nutrients from aquatic vegetation to the Coeur d Alene Lake system was higher from the Lower Lakes area than from Coeur d Alene proper, indicating the importance of these shallow water habitats. The overall conclusion offered from this baseline assessment of submersed aquatic vegetation in the Lower Lakes area of Coeur d Alene Lake is that this growth is healthy, very productive and reasonably diverse. The plants that were identified in the Lower Lakes transects and grid point sampling were all native species with the exception of Myriophyllum spicatum (Eurasian watermilfoil) which was found widely distributed throughout this area with limited dense growth areas in Chatcolet and Round Lakes. It is expected that this presence will increase significantly in the coming years, absent implementation of control measures. However, harvesting of aquatic vegetation as a means of controlling nutrient inputs to the lake must be further evaluated to determine its cost effectiveness. Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 3

9 Acknowledgements The Coeur d Alene Tribe s Lake Management Department wishes to acknowledge and thank the Basin Environmental Improvement Project Commission (BEIPC) for funding this project as part of their Clean Water Act grant program. Aquatic vegetation has been largely overlooked in past investigations of Coeur d Alene Lake and the Lower Lakes area (Chatcolet, Round and Benewah Lakes) even though it appears to contribute significant levels of nutrients to the larger lake system and likely affects the overall quality of this resource. The Lower Lakes area, in particular, provides the most extensive and fertile submersed plant growth area in this system. The Lake Management Department also wishes to acknowledge the Coeur d Alene Tribal Council and their willingness to undertake this project. Just as past investigations have overlooked aquatic vegetation, this arena of lake management is not a topic for light discussion. Therefore, the Council had to give this their full attention to be able to understand the importance of aquatic vegetation and the significance of aquatic weeds such as Eurasian watermilfoil. And finally, we wish to acknowledge Avista Corporation for funding the Tribe to perform a parallel study of aquatic vegetation throughout Coeur d Alene Lake. This parallel study allowed us to develop our aquatic plant nutrient loading estimate from a much larger data base which resulted, we feel, in a much more comprehensive and useful calculation for Coeur d Alene Lake as a whole. David S. Lamb Lake Ecologist Project Lead Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 4

10 Introduction The development of a Lake Management Plan for Coeur d Alene Lake was initiated in 1991 in response to long-term concerns by the Tribe and others over water quality degradation in the lake. These concerns centered on increases in water column nutrient levels and heavy metals contamination found in lakebed sediments. The Tribe, State of Idaho and US Environmental Protection Agency (USEPA) have determined that management of metals-contaminated sediments could best be accomplished through effective nutrient management. This determination is supported by the Basin-wide Remedial Investigation (US EPA 2001a) and Feasibility Study (US EPA 2001b). The Tribe and States of Idaho and Washington concurred with the Remedial Investigation / Feasibility Study and a formal Record of Decision was signed in September 2002 (US EPA 2002). Nutrient management has been a key concern of the Coeur d Alene Tribe for many years. This is due to the increase in productivity and concurrent decrease in water quality that excessive levels of phosphorus and nitrogen can cause. Increased productivity can directly result in lower hypolimnetic oxygen levels which can, in turn, cause the release of toxic heavy metals which have accumulated in the sediments of Coeur d Alene Lake. The cleanup of metals contamination in soils and waters within the Coeur d Alene River and Lake system is one of the primary goals of the Basin Environmental Improvement Project Commission (BEIPC). A related issue in the Coeur d'alene Lake area is the potential for infestation by the noxious, invasive plant Eurasian watermilfoil (Myriophyllum spicatum). This plant has been found in most of the nearby lakes in Idaho and Washington (Hayden, Spirit, Pend Oreille, Liberty and Newman Lakes in particular) and is known to be introduced into un-infested waters via fragments which can be carried on boats and boat trailers. This plant represents a significant threat to lake beneficial uses including fish and wildlife habitats. Once this plant becomes established it is virtually impossible to eradicate and with continued spreading, control can be routine and costly. The 2002 Clean Water Act appropriation for the US EPA included a line item for a grant to the State of Idaho (BEIPC) to undertake activities that conduct and promote the coordination and acceleration of research, investigations, experiments, training, demonstrations, surveys and studies relating to the causes, effects, extent, prevention, reduction and elimination of pollution [Clean Water Act, Section 104 (b)(3)]. Thus, a Coeur d Alene Lake aquatic vegetation survey project, including an effort to estimate nutrient release from existing aquatic plant populations, was considered appropriate for funding through the BEIPC research and demonstration project program. The proposed project is warranted by the following needs: need for lake-wide biomass and distribution data on which to base harvest or other control treatment plans and effectiveness assessments, need for species-specific nutrient content data to develop estimates of nutrient release (loading) to Coeur D Alene Lake (this also would effect harvest plans), and need for surveillance to document the presence or absence of invasive, noxious aquatic species such as Eurasian watermilfoil. Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 5

11 The southern end of Coeur d Alene Lake, an area referred to collectively as the Lower Lakes, and more specifically as Chatcolet, Round and Benewah Lakes, was chosen for the initial (pilot) phase of this work because of the presence of extensive shallow water habitats with nutrient rich sediment substrates; i.e. prime aquatic plant growth areas. Following submission of a project proposal to the BEIPC Leadership Group review process and approval by the BEIPC Board, funding through the Clean Water Act grant program was approved and an Idaho Department of Environmental Quality (IDEQ) Sub-Grant Agreement was issued June 1, This report documents the findings of this Tribal project. It is hoped that these findings will allow clarification of certain questions related to the nutrient loading to the lake from submersed aquatic plant growth. The results of this study have been evaluated in light of those from a parallel study of aquatic vegetation in the larger Coeur d Alene Lake area, funded by Avista Utilities, to provide an overall assessment of the impact of submersed plant growth on the nutrient budget of Coeur d Alene Lake, an important Tribal and regional resource. Description of Study Area The three lakes designated for this aquatic vegetation study surround the lower St. Joe River which enters the south end of Coeur d'alene Lake (see Figures 1 and 2). Starting upstream of the town of St. Maries, the St. Joe was historically a low-gradient, sinuous system passing through a wide valley of extensive small lakes and wetlands. Historically, natural levees lined the river channel out into Coeur d'alene Lake proper. In 1906 the Post Falls Dam was built on the Coeur d Alene Lake outlet (the Spokane River, see Figure 1) which raised the surface elevation of this lake and river system and allowed the control of the summer pool levels at 2,128 feet; approximately eight feet above the historical levels at that time of year. These elevated pool levels caused the inundation of the valley area surrounding the lower St. Joe River, forming a series of shallow "chain lakes" which were subsequently named Benewah, Chatcolet, Hidden and Round Lakes (see Figure 2). The channel of the St. Joe River is lined with the remnants of the former levees although there are breaks in these which allow river water to flow into the chain lakes at numerous locations. Along the lower two miles of the St. Joe these historic natural levees have been reduced through erosion over the last 100 years to mere submerged berms. The surface area of Coeur d Alene Lake at summer full pool is approximately 32,000 acres (Clean Lakes Coordinating Council, 1996). The approximate areas of the three lakes included in this study, again at full pool, are 1,700 acres for Chatcolet, 1,000 for Round and 500 for Benewah (estimated from CDA Tribe GIS map data; actual lake boundaries are not well defined). While the maximum depth of Coeur d Alene Lake is over 200 feet and the mean depth is 72 feet (CLCC 1996), the maximum depth in the Lower Lakes is only about 35 feet in Chatcolet (Northwest Map Service 1996) and the mean depth is likely between 8 and 10 feet (rough estimate by Tribal staff). Until recently the bathymetry of the Lower Lakes was not known with any certainty with only a few spot depth soundings recorded. In 2003, however, a bathymetric survey was performed for Avista Corporation s Post Falls Dam relicensing effort. Depth information from that effort was corrected by Tribal GIS Program staff and selected bathymetric contours appear on Figure 2. Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 6

12 Figure 1. Coeur d'alene Lake, Idaho, showing Lower Lakes Aquatic Vegetation Study Area. Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 7

13 Figure 2. Map of Coeur d Alene Tribe s Lower Lakes Aquatic Vegetation Survey study area showing grid points and transect locations. Purpose / Objectives Study Purpose The primary purpose of this study is to develop baseline data on submersed aquatic plant species distribution and biomass in Benewah, Chatcolet and Round Lakes. The secondary purpose is to estimate nutrient (phosphorus and nitrogen) release from the existing plant beds into the water column of these lakes and, subsequently into Coeur d'alene Lake. The tertiary purpose is to inspect these lakes for the presence of invasive, noxious aquatic species. Study Objectives Specific objectives are to perform SCUBA diver collection of submersed aquatic vegetation species along set transects, and to perform additional sampling at randomly selected grid Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 8

14 intersections using remote (weed rake) sample collection techniques. All diver-collected samples were sorted by species and each species sample was dried to obtain standard biomass measurements. Sub-samples of diver-collected plants were analyzed to determine nutrient (phosphorus and nitrogen) content. Samples collected at grid intersection sites were sorted to identify species present and discarded. Based on the biomass and nutrient content data, and published literature on species-specific nutrient release rates, estimated nutrient loading from submersed plants was calculated. Materials and Methods Transect Sample Collection Techniques The transect surveys were performed during the mid-july to mid-august period in 2004 and The quantitative sampling was a modification of the "line intercept" method (APHA, 1995). Samples were collected using SCUBA techniques along fixed lines (transects) which were oriented from a start point on shore by a compass heading. The sampling crew consisted of two certified divers and a boat operator. Typically, only one diver was in the water at a time and the other helped bag and log samples. The boat used for this project was a twin outboard, 24 ft Munson PackCat landing craft (see Project photograph #4 in Appendix C). There were 24 transects sampled as shown on Figure 2. Along these transects samples were collected at one-foot depth intervals using a "quadrat"; a fixed-corner, three-sided frame that defined a standard sampling area (18 inches square which equals 2.25 square feet or square meters). The quadrat is shown in Figure 3 and also Photographs #3 - #5 in Appendix C. At each designated sampling location along the transect line, the quadrat was placed on the lake bottom under and around any plants present and said plants were collected by hand and placed in a mesh bag carried by the diver. The mesh bag was then taken to the surface and given to a boat attendant who would rinse the sample by dipping the mesh bag in the lake water until the rinsate appeared clear, spin the bag to shake out excess water, transfer the sample to a plastic bag labeled with the date, transect number and depth and place each sample bag on ice in a cooler. Figure 3. Quadrat and mesh bag used for aquatic vegetation sampling portion of Lower Lakes Aquatic Vegetation Survey project. Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 9

15 Because the diver s depth gauges read in even feet of depth, the samples were actually collected between feet, feet, feet, etc. of depth. As the diver descended along the lake bottom the samples would be collected as soon as the depth gauge read the even foot, therefore closer to the desired one-foot interval than if the diver was ascending. In addition, if the lake bottom along the transect was nearly flat for a long distance (over 150 feet, approximately), or rose and fell so that the one-foot intervals were passed repeatedly, additional samples were collected. These additional samples were labeled with the date-transect-depth and a letter b, c, d etc to designate the successive samples. One of the observations that the divers were asked to make during the sample collection work was that of sediment materials present at the sample sites. While the expected predominant sediment type was organic muck, no formal texture class system was set up prior to the sampling work. Afterwards, however, recorded sediment descriptions were categorized in the following classes, which follow standard USDA textural descriptions (Virginia Department of Health, 2005): Rocky -- greater than 150 mm (six inch) diameter particles Stony mm to 150 mm diameter particles Gravelly -- 2 mm to 10 mm particles Sandy mm to 2 mm particles Silt mm to 0.05 mm particles (this includes organic muck materials), and Clay -- less than mm particles (this often took the form of hardpan). Transect Sample Sorting Techniques At an on-shore location, typically at the Tribe s Natural Resources / Lake Management office building in Plummer, ID, all collected plant materials were sorted by project staff within 24 hours of collection (see Photograph #6 in Appendix C). Each bagged sample was placed in a 16 inch by 24 inch by 6 inch deep plastic pan and spread out to separate the biomass as much as possible. As plants were identified they were removed by hand and placed into paper sacks labeled with the date, transect, depth and a three-letter species identifier. Plant samples were sorted to species, whenever possible, or to genus. Two sizes of paper sacks were used (Weyerhaeuser #8 and #420) to accommodate larger or smaller volumes of the sub-samples. Following the sorting, the paper sacks containing the sorted sub-samples were folded over and stapled to prevent loss of plant material during transport. Bagged sub-samples were placed in topless cardboard boxes and allowed to air dry for between three and five days before delivery to the contract laboratory. This prevented failure of the sacks because of the wetness of the samples (especially large ones) and also may have prevented the growth of mold in the subsamples prior to laboratory analyses. Grid Sampling Methods Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 10

16 For this effort a grid with 250 meter intervals each way was set up across the lower three lakes using GIS mapping techniques. Grid intersections (nodes) that fell on land were deleted and numerical identifiers were designated for the remaining nodes. A list of coordinates for all sites (nodes) was developed as part of the data collection form. A hand-held GPS unit (Garmin E- Trex Legend) was used to navigate to each site to be sampled (typically within a four to eight meter radius of the point) and one toss was made with a weighted weed rake (see Figure 4). The rake was retrieved and species present on the rake were recorded. This sampling was performed following the transect surveying (i.e. early-september) in 2004 and Data from the grid sites was tabulated and is described herein (see below). Figure 4. Weed rake used for grid sampling portion of Lower Lakes Aquatic Vegetation Survey project. Laboratory Analyses All sorted sub-samples were delivered to Spokane Tribal Labs in Spokane, WA, the selected analytical facility for this project. Requested analyses were dry-weight biomass and nitrogen and phosphorus content. All sub-samples and the requested analyses were recorded on Chain-of- Custody forms which were kept on file by both the lab and the Tribe s Project Lead. Each Chain-of-Custody form had space to list only 10 sub-samples so the sub-samples for each form were placed together in a grocery sack on which was written a unique identifier which also appeared on the Chain-of-Custody form. Sub-samples to be analyzed for nutrients were designated by the Tribal Project Lead following sorting of collected samples. Sub-samples were chosen for analyses based on there being a large enough volume of plant material for the analyses (thus these were the larger sub-samples) and on the approximate proportion of the number of sub-samples of each species. Approximately 10% of the sub-samples were designated for nutrient analyses. Biomass Analyses Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 11

17 The contract laboratory used a modification of Standard Method #10400 D. 3. a. (APHA 1995) to determine Biomass. This involved drying each plant sample sack in a forced-air oven at 105ºC for hours, cooling in a desiccator and weighting. Approximately 10% of the subsamples were dried a second time, cooled and re-weighed as part of the Quality Control procedures (see Appendix B). A number of empty paper sacks were also dried in this manner and weighted so that the sack weight could be subtracted from the sample-plus-sack weight. The results of this are also presented in Appendix B. The laboratory reported weights in grams (g) and biomass was calculated by Tribal staff based on the area of the quadrat to yield grams per square meter (g/m 2 ) biomass. Total Phosphorus Content The contract lab followed EPA Method to process and analyze collected, dried plant material for phosphorus content. The sub-samples to be analyzed were ground to powder using an Ultimate Chopper Model CH-1 food processor. Phosphorus was reported as milligrams phosphorus per dry kilogram of plant (mg/kg). Total Nitrogen Content The contract lab followed EPA Methods and to process and analyze collected, dried plant material for Total Kjeldahl nitrogen (TKN), nitrate (NO 3 ) and nitrite (NO 2 ) nitrogen content. TKN is a measure of nitrogen in organic compounds (i.e. plant cell materials) and as ammonia (NH 3 )(APHA 1995). Each nitrogen fraction was reported as micrograms nitrogen per dry kilogram plant (mg/kg). Nutrient Loading Calculations Nutrients are known to be important in assessing and monitoring water quality, and aquatic vegetation has been shown, in some instances, to influence lake water quality by drawing nutrients (most often phosphorus) from the sediments and recycling them into the open water (Wetzel 2001). Therefore, one of the goals of this project was to estimate potential nutrient (phosphorus and nitrogen) release from the existing plant beds into the water column of the project area. To do this the baseline survey of species presence, biomass and nutrient contents had to be completed and a literature review had to be performed to obtain information on the processes and rates that nutrients would (or could) be released from the aquatic plant species that are present. The literature review that was performed included a search for published scientific documents related to nutrient release, regeneration or recycling by aquatic plant species. While an initial effort was undertaken utilizing the Internet search engine Google, most of the effort was focused on searches of natural science databases available through Eastern Washington University and Washington State University. Communications with regional experts in lake management also resulted in a number of document findings. Digital or paper copies of Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 12

18 references were obtained whenever possible and these were closely studied for pertinent information. While this literature search cannot be considered exhaustive, it did produce a number of helpful documents, most of which were published during the 1970s and 1980s. Further search and procurement of additional (especially more recent) articles would likely allow some refinement of the numerical coefficients used in the nutrient loading estimation. With the completion of the literature search, aquatic plant species biomass and nutrient data were compiled based on the studied transects. Transect data was expanded to represent a number of Aquatic Vegetation Regions in the lakes and the area of each of these regions was calculated based on the depths that submersed plants were found along the representative transect(s) for each area. With the area determination (in square meters, m 2 ), a calculation of project-wide average biomass for each species (in grams dry weight per square meter, g/m 2 ) and project-wide nutrient content for each of the predominant species (in micrograms phosphorus or nitrogen per gram of dry weight biomass) the project-wide total nutrient content for each species was calculated. This was then applied to the nutrient release criteria obtained from the literature search and the estimated annual release of phosphorus and nitrogen were calculated. The results of this analysis are presented in the Discussion section, below. Quality Assurance and Quality Control Quality assurance and quality control (QA/QC) procedures for this project followed the Quality Assurance Project Plan (QAPP) which was prepared for this project and approved by the Idaho Department of Environmental Quality and the US EPA. QA and QC results specific to this project are summarized in Appendix B. The quality objective of the Lower Lakes Aquatic Vegetation Survey project, is to obtain representative data on aquatic species presence, biomass and nutrient (nitrogen and phosphorus) content. To this end, quality criteria focus on field sampling and sample sorting, laboratory analyses and data interpretation. It is important to note that aquatic vegetation growth and distribution is influenced by many factors, including light availability (water clarity and shading are co-factors), water chemistry, sediment texture and composition, depth, slope, disturbances and presence of herbivores. The result of variations in these factors is an often patchy, irregular spatial distribution with varying plant densities of often intermixed species. The presence of invasive species, however, can result in extensive areas being occupied exclusively by the invader. With poor water clarity or dense plant growth, it is often difficult for someone conducting aquatic vegetation sampling to determine if they are collecting representative samples. Therefore quality criteria for this project focused on the collection of as many samples as possible along with detailed information on depth, location and sediment conditions at the sample sites. The criterion for field sample collection is to obtain unbiased samples at multiple depths along multiple transects through the study area. Transects were laid out to make maximum use of the Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 13

19 project s resources and cover as much of the lake bottom as possible. Divers were trained to closely follow the compass heading of the transect and to find successive sampling depths without regard for the plant species that might be present. The criterion for sample processing was the systematic sorting of collected samples by logical, observable physical characteristics. Physical characteristics allowed separation of collected samples into individual species or into similar groups of species and each of these became a subsample to be submitted for laboratory analysis. Plant species identification followed nomenclature presented in a regional plant manual; in this case An Aquatic Plant Identification Manual for Washington s Freshwater Plants (Washington Department of Ecology 2001). The sorting process retained all plant shoots and leaves but removed all non-plant or non-living material as well as plant roots, rhizomes or other structures which are found below the sediment surface. Sub-samples were placed in bags clearly marked with the appropriate species identifier, the date collected, the transect number and depth. The criterion for laboratory analyses was the accurate determination of desired plant physical and chemical characteristics: dry weigh and nutrient content. This was accomplished by using an accredited laboratory which followed an approved Quality Assurance Plan. Analytical detection limits, precision and accuracy levels (referred to as Data Quality Indicators) appropriate to this project are presented in Table 1. Table 1. Data Quality Indicators applicable to the Coeur d Alene Lake Baseline Aquatic Vegetation Survey project. Parameter Instrument Reporting Units Detection Limits Precision Accuracy Reference Analytical Method EPA/Standard Methods Biomass Gravimetric g 0.01 g +/- 20% +/- 20% SM10400 D Total Semi-automated mg/kg 2.50 mg/kg +/- 20% +/- 25% EPA Phosphate Colorimetry Nitrate Ion Chromatography mg/kg 1.00 mg/kg +/- 20% +/- 10% EPA Total Kjeldahl Nitrogen Semi-automated Colorimetry mg/kg 25.0 mg/kg +/- 20% +/- 25% EPA The criteria for data interpretation were to use logical organization to tabulate plant biomass data in relation to location, depth, species and nutrient content, and to use accepted statistical analyses to illuminate significant similarities which describe the growth of submersed vegetation across the study area. Because of the typically patchy distribution of aquatic vegetation in the littoral zone of lakes, and the need for destructive sampling to actually remove vegetation from the Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 14

20 sample sites for biomass determination, there were few general Quality Control procedures that applied to this project. As indicated, the quality objective of this project is to obtain representative data on aquatic species presence, biomass and nutrient content. Therefore, quality control in the field focused on sampling and sorting while office procedures focused on data handling. Results General Throughout this report the terms macrophytes, aquatic vegetation and aquatic plants are used interchangeably. What are referred to by each of these terms are those plants which are macroscopic vascular angiosperms (plants having their seeds enclosed in an ovary, as opposed to gymnosperms and algae) and completely submersed, although some may have flowering structures which extend above the water surface. While some forms of macroscopic algae are found in the study area (i.e. Chara and Nitella), and when found in our samples these were submitted to biomass analysis, these are not a focus of this study and this data is not included in the summaries herein. With -19.a/.a/. >- solrojeieldr(of this pin )]T004 Tc T Troj, many/. >-ss anaes a opsib1(ied showiv

21 Bearing Length Depth Range Primary Transect # Start Point Coordinates* (degrees) (ft) of Aq. Veg. ** Substrate C , muck C rocky C , muck C muck C muck C , muck C , muck C muck C , muck C , muck C , muck C , clay & muck R , muck R , muck R , muck R , muck R , muck R , muck R , muck B muck B muck B muck B muck B , muck * UTM Zone 11N, NAD 1983 ** Depth range where plants found along transect: beginning - maximum (or minimum) - end, all in feet below normal summer pool elevation of 2,128 feet.

22 Regions used for the Estimate of Nutrient Loading effort, Chatcolet Lake is further subdivided into three areas based on observed bottom substrate type, as described below.) Table 3. Species codes and scientific and common names of plants sampled during the Lower Lakes Aquatic Vegetation Survey project. Species Code Cha CD Esp Msp Iso Nit Nsp PA PE PF PR PRi Psp PZ RA Ssp UV UNK UNKP Full species name (common name) Chara species (macro-algae) Ceratophyllum demersum (coontail) Elodea species (water weed) Myriophyllum species (milfoil) Isoetes species (quillworts) Nitella species (macroalgae) Najas species (water-nymph or niad) Poptamogeton amplifolius (big-leaf pondweed) Potamogeton epihydrous (ribbon leaf pondweed) Potamogeton friesii (flat-stalked pondweed) Potamogeton robbinsii (fern-leaf pondweed) Potamogeton richardsonii (Richardson's pondweed) Potamogeton species (thin leafed pondweeds) Potamogeton zosteriformis (flat-stem pondweed) Ranunculus aquatilis (white water buttercup) Sagitaria species (submersed wapato?) Utricularia vulgaris (common bladderwort) Unknown species (not identifiable) Unknown pondweed species Grid Point Results The results of the grid sampling for the two years of this project are presented in Appendix B. In 2004, 197 grid points were visited. Of these grid points 52 had no vegetation (primarily because the depth was too great) and 34 had primarily emergent plants so no sample was collected. Of the remaining grid points, the average number of species seen per sample (based on when any submersed species or species groups were found) was 2.9. The top three species groups, in terms of the number of samples they were found in were: Esp (101 samples), Msp (93 samples) and Psp (66 samples). In 2005, 194 grid points were visited, 53 had no plants and 31 were dominated by emergent species. The average number of submersed species was 2.5 and the top three species groups were, again Esp (90 samples), Msp (64 samples) and Psp (50 samples). Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 17

23 While not related specifically to submersed plant community structure, Figure 5 below indicates the results of the grid point sampling insofar as where submersed plants were found, where emergent plants were found and where there were no plants found. Figure 5. Location of submersed and emergent plant types, as well as areas where no plants were found, from the grid point sampling, Lower Lakes Aquatic Vegetation Survey Project. Transect Results In Benewah Lake in 2004 the average number of species per sample was 3.3 and this ranged from one to seven species. The most predominant species in terms of frequency of occurrence was Elodea canadensis (Esp) but Potamogeton robinsii (PR) and then Myriophyllum spicatum (Msp) and the thin leafed Potamogeton species (Psp) were also frequently encountered. In 2005, Benewah Lake was only sampled along four of the six transects sampled in 2004 and the resultant average number of species per sample was 2.9, ranging from one to six species. In these four transects (B2, B3, B4 and B6) the predominant species, again in terms of frequency of occurrence, was PR with Esp and Msp as secondary species. It is notable that PR was seldom seen in transects from Chatcolet or Round Lakes. Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 18

24 In Chatcolet Lake the average number of species per sample was 3.0 in 2004 and 3.2 in Of the ten transects sampled in 2004 one was dropped (C6) and tw

25 Overall Average Standard Number of Minimum Maximum Species Transects found at Biomass (g/m2) deviation (s) samples (n) (g/m2) * (g/m2) B2., B3, B5, B6, C1-C12, R1, CD R4-R Esp B1-B6, C1-C8, C10-C12, R1, R4-R Iso B2, B3, B5, B6, C1-C8, C10- C12, R1, R4-R Msp B1, B2, B4, B6, C1-C12, R1, R4-R Nsp B2-B4, C1, C3-C8, R1, R4-R PA B n/a PE B n/a PF C n/a PR B2-B4, C2, C5, C PRi B2-B4, B6, C1, C3-C7, C9, C11-C12, R1-R3, R5-R7, Psp B1-B6, C1-C12, R1-R PZ C1, C2, C5, C RA B1, B3, B6, C1, C4, C5, C7, C9-C12, R1, R2, R Ssp B2, B4, B6, C3-C6, C9-11, R1, R2, R UV B * 0.01 is the arbritrary value assigned when the lab reported a ND (non-detect) for sub-samples. The species found at the highest individual sub-sample biomass were Ceratophyllum demersum (CD, 1, g/m 2 ), Elodea species (Esp, g/m 2 ), thin-leafed Potamogeton (Psp, g/m 2 ) and Myriophyllum spicatum (Msp, g/m 2 ). These levels of biomass are atypical, however, as indicated by the mean biomass values for these species of 30.37, 49.94, and g/m 2, respectively. The highest mean biomass seen was for PR at g/m 2. The biomass data compiled by transect (all species and depths) is presented in Table 5 and Figures 6 and 7. As shown in Figure 6, the highest average sample biomass was g/m 2 seen at Transect B5 although this transect also had the smallest number of samples of any of the transects. Interestingly, the second highest average biomass, g/m 2 was from Transect C7 which had the highest number of samples. The lowest per sample biomass of g/m 2 was found on Transect C2. Again, the high standard deviations calculated indicate the high degree of variability in this summarization. In order to show what the variation in aquatic plant biomass between the two years of this project, average sample biomass by transect was determined for both study years. The result of this analysis is presented in Figure 7. As stated at the beginning of this section, above, it is felt that the two years sampling work should not be considered to be repeats of the same transect because of the variations in actual compass course followed by the divers and to variability in Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 20

26 Table 5. Summarized aquatic vegetation biomass by transect for the Lower Lakes Aquatic Vegetation Survey project. Overall Average Standard Number of Minimum Maximum Transect Biomass (g/m2)* deviation (s) samples (n) (g/m2)** (g/m2) B B B B B B C C C C C C C C C C C C R R R R R R R * Per sample biomass, not per transect ** 0.01 is the arbritrary value assigned when the lab reported a ND (non-detect) for sub-samples. locating the GPS start point for the transect. With that in mind, however, many of the transects, especially those with average sample biomass levels less than 40 g/m 2 or so were very close. It was the transects with the highest per-sample biomass levels that differed the most. Figure 7 also shows which transects were sampled during only one of the two years. Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 21

27 Biomass (g/m2) Mean Sample Biomass by Transect (all species except Cha, Nit and Ssp) B1 B2 B3 B4 B5 B6 C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 R1 R2 R3 R4 R5 R6 R7 Transect Figure 6. Average aquatic vegetation biomass variations (all species) by transect for the Lower Lakes Aquatic Vegetation Survey project. Average Biomass by Transect and Year Biomass (g/m2) B1 B2 B3 B4 B5 B6 C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 R1 R2 R3 R4 R5 R6 R7 Transect Figure 7. Average aquatic vegetation biomass variations (all species) by transect and year for the Lower Lakes Aquatic Vegetation Survey project. Aquatic vegetation summarized by depth (across all transects and species) is shown in Table 6 and Figure 8. These indicate that submersed plants were found at depths between 1 and 18 feet, although there were few sites that had any vegetation at the 16 or 18 foot depths. There is an apparent progression in biomass between the 2 and 12 foot depths, with the highest sample biomass of g/m 2 at 12 feet, and then a rapid decline to the 18 foot depth. The reason for the difference between the 1 and 2 foot depths is likely due to the small number of samples at the Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 22

28 Table 6. Summarized aquatic vegetation biomass by depth for all transects sampled during the Lower Lakes Aquatic Vegetation Survey project. Overall Average Standard Number of Minimum Maximum Depth Biomass (g/m2)* deviation (s) samples (n) (g/m2)** (g/m2) * Per sample biomass, not per transect ** 0.01 is the arbritrary value assigned when the lab reported a ND (non-detect) for sub-samples. 1 foot depth because field observations pointed to there being very little vegetation at the 1 and 2 foot depths, presumably due to wave turbulence near the shore. The maximum sample biomass was found at the 9 foot depth and this was the 1, g/m 2 Ceratophyllum demersum sample described above. The next two highest biomass results were at the 11 and 12 foot depths, 1, and 1, g/m 2 respectively, and these were mixed species samples. The minimum biomass at most depths was the 0.01 g/m 2 which indicates that the dried sample was less than the weight of the paper sacks the samples were placed in. Nutrient Results All aquatic vegetation nutrient content results are presented in Appendix A. Nutrient analyses were only performed on 13 species collected from the sampled transects. These species were Ceratophyllum demersum (code CD), Elodea species (Esp), Myriophyllum species (Msp), Nitella macro-algae (Nit), Najas species (Nsp), Potamogeton amplifolius (PA), P. epihydrous (PE), P. Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 23

29 Mean Biomass (g/m2) Mean Sample Biomass by Depth (all species except Cha, Nit and Ssp) Depth(ft) Figure 8. Aquatic vegetation biomass variations by depth for all transects sampled for the Lower Lakes Aquatic Vegetation Survey project. robinsii (PR), P. richardsonii (PRi), the thin-leafed Potamogeton species (Psp), Ranunculus aquatilis (RA) and Sagitaria species (Ssp). Tables 7 through 9 present the averages, standard deviations, maximum and minimum values of the three nutrient analyses that were performed on the sub-samples. These analyses were total phosphorus (TP), total Kjeldahl nitrogen (TKN), nitrate nitrogen (NO 3 ) and nitrite nitrogen (NO 2 ), as described in the Methods section, above. The overall average and standard deviation, that is of all sub-samples of each species (regardless of transect or depth) is the only summarization performed on this data. This is due to there being too few nutrient results to assess nutrient content by transect or depth. However, it can be seen in Tables 7 through 9 that the variability is considerably less than that of the biomass data. Nitrite data were below detection in all samples analyzed and are not further discussed. The TP summarization shown in Table 7 indicates that Sagitaria species (Ssp) contained the highest concentration of phosphorus, with a mean of 7,611 μg P/g and a maximum of 9,320 μg P/g. However, this species is not considered to be a true submerged species and contributes little to the overall submersed plant biomass. Of the six predominant species that were found during this study (CD, Esp, Msp, PR, PRi and Psp), the highest mean TP concentrations were 4,084 and 4,055 μg P/g, with Esp and CD, respectively. The lowest mean phosphorus concentration of the plants analyzed was 2,709 μg P/g in PRi. The organic forms of nitrogen (including ammonium), as measured by the Total Kjeldahl Nitrogen (TKN) analysis, were by far the predominant forms of nitrogen in all of the vegetation samples (Table 8). The highest mean TKN (30,167 μg N/g) was seen with Ssp and the single maximum value was 42,200 μg N/g with Esp. The lowest mean TKN was 22,176 with PRi and Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 24

30 Table 7. Total phosphorus (TP) data collected for the Lower Lakes Aquatic Vegetation Survey project. Overall Average Standard Number of Minimum Maximum Species TP (μg/g) deviation (μg/g) samples (n) TP (μg/g) TP (μg/g) CD 4,055 1, ,310 8,140 Esp 4,084 1, ,990 Msp 3, ,650 5,230 Nit 3, ,480 3,270 Nsp 4,047 1, ,490 5,810 PA 3,560 n/a 1 3,560 3,560 PE 4,010 n/a 1 4,010 4,010 PR 3, ,450 3,830 PRi 2, ,000 Psp 3,047 1, ,740 RA 4, ,820 5,360 Ssp 7,611 1, ,280 9,320 Note: μg/g = mg/kg the lowest absolute value was 4,710 with Psp. However, in spite of these variations, the similarity in TKN results across the species tested is worthy of note. Again, looking at the six primary species (CD, Esp, Msp, PR, PRi and Psp) the range of mean TKN values was only between 22,176 and 27,757 μg N/g and the standard deviation values were substantially below the mean values, an indication of lower variations in the individual analyses. The nitrate (NO 3 ) concentration data, while only a small fraction of the TKN concentrations, were somewhat variable across the species tested (Table 9). The highest mean was 48 μg N/g for PR and the lowest means were 3 μg N/g for Nit, the macro-algae, and 8 μg N/g for both PE and Psp. The maximum nitrate concentration was 192 μg N/g in a Msp sub-sample and the minimum was 0.1 μg N/g in a Psp sub-sample. Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 25

31 Table 8. Total Kjeldahl nitrogen (TKN) data collected for the 2005 Baseline Coeur d'alene Lake Aquatic Vegetation Survey project. Overall Average Standard Number of Minimum Maximum Species TKN (μg/g) deviation (μg/g) samples (n) TKN (μg/g) TKN (μg/g) CD 25,360 5, ,500 38,100 Esp 23,884 5, ,550 42,200 Msp 27,283 6, ,300 40,600 Nit 27,925 9, ,200 41,700 Nsp 26,167 8, ,800 34,100 PA 30,700 n/a 1 30,700 30,700 PE 26,200 n/a 1 26,200 26,200 PR 27,757 2, ,600 33,700 PRi 22,176 3, ,600 29,800 Psp 24,608 6, ,710 34,300 RA 24, ,900 25,500 Ssp 30,167 3, ,000 35,200 Note: μg/g = mg/kg Table 9. Nitrate nitrogen (NO 3 ) data collected for the 2005 Baseline Coeur d'alene Lake Aquatic Vegetation Survey project. Overall Average Standard Number of Minimum Maximum Species NO 3 (μg/g) deviation (μg/g) samples (n) NO 3 (μg/g) NO 3 (μg/g) CD Esp Msp Nit Nsp PA 10 n/a PE 8 n/a PR PRi Psp RA 16 n/a Ssp Note: μg/g = mg/kg Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 26

32 QC Results Duplicates Biomass Duplicates Biomass QC results are presented in Table C1 in Appendix C in the form of duplicate tests. A total of 122 sub-samples (8% of the total number of sub-samples analyzed) were re-dried, cooled and re-weighed to test the precision of the drying and weighing process. The comparison of initial and duplicate results is presented as the Relative Percent Difference (RPD). From the lab s QAP, RPD is calculated as follows: (original result duplicate result) RPD = X 100 (original result + duplicate result)/2 The confidence limit (RPDCL) for this is 20% and all but 10 results were well within this limit. In fact the mean RPD of those results that were within the RPDCL was 5.39 with a standard deviation of The 10 samples that were not within the RPDCL were either not controlled by the RPDCL value because the sample results were less than ten times the Reporting Limit (0.1 gram, i.e. very low biomass levels) as indicated by note QR-01 in Table B1, or both percent recoveries were acceptable and sample results for the QC batch were accepted based on percent recoveries and completeness of QC data, as indicated by note QR-02 in Table B1. Nutrient Duplicates Duplicate nutrient QC results are presented in Tables C2, C3 and C4 in Appendix C. Seventeen samples for each of the three analyses [Total Phosphorus (TP), Total Kjeldahl Nitrogen (TKN) and Nitrate) were duplicated in order to test the precision of the nutrient analyses. Again, Relative Percent Difference was determined from the difference between initial and duplicate results. In the case of TP and TKN, which have a RPDCL of 25%, all duplicate results were within this limit. Specific mean ± standard deviation of RPD values were 2.40 ± 1.25 for TP (actual range ) and 3.31 ± 3.07 for TKN (actual range ). The nitrate RDPCL is 20% and all but one duplicate result were within this limit. The RPD mean value for nitrate was 8.56 ± 7.85, with an actual range of Nutrient Matrix Spikes To test the accuracy of the analyses, nutrient spikes of known concentration were added to certain sample digestates and analyses were re-run (see Tables C5, C6 and C7). For TP the spikes ranged from 250 to 2,000 mg/kg, for TKN they were 1,000, to 5,000 mg/kg and for nitrate the spikes were 10 to 1,000 mg/kg. QC results were based on the percent recovery of the combined sample plus spike and the applicable confidence limits are % of the expected result. Actual percent recovery determined for this analysis was 71% 124% for the TP testing, 60% 128% for TKN and % for nitrate. Thus there were samples for each analysis that were outside the confidence limits. However, based on the explanatory notes provided with the laboratory data, all matrix spike results were deemed acceptable. Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 27

33 Nutrient Reference Samples Nutrient reference analysis results are presented in Table C8. Analysis of reference materials of known nutrient concentrations was performed to provide an important indication of method accuracy. In this case analytical standard apple leaves were purchased and put through the various nutrient tests. Analytical results were compared to the reference (known) concentration and a percent recovery was determined. Overall the recovery percent ranged from 86.2% - 114% for TP, TKN and nitrate. The applicable confidence limits were between 75% and 125% for TP and TKN and between 77.4% and 122.6% for nitrate. Thus, all tests were within the prescribed confidence limits. Nutrient Blanks Nutrient blank analysis results are presented in Table C9. The contract laboratory checked blank samples for the presence of contaminants in the dilution water and reagents. All blank tests came back below detection so it can be said that there was no contamination above the detection limits. Discussion Grid Results Versus Transect Results The difference in number of species (or species groups) between the grid and transect sampling is undoubtedly due to the increased rigor of the transect sampling where the samples were collected by the divers so as to obtain all of the plant material present within the quadrat and then the samples were sorted in a pan so as to obtain the smallest identifiable portion of all plants present. By comparison, the weed rake tosses undoubtedly missed plants growing close to the bottom under dense canopy and simply pulling the collected sample apart also could lead to missed species. Regarding the biomass results, the variability of this data is high, as indicated by the high standard deviation values presented (with few exceptions the standard deviation is one to two times the mean value). This is undoubtedly influenced by the fact that the plants are found at locations with differing depths, slopes, aspects and substrates and also due to the patchy distribution noted by the divers. Comparison of Project Results with Those of Other Studies Biomass Since the studies reviewed in the literature search indicated that nutrient release information were often conducted in-situ in lakes, or using vegetation collected from lakes, it is possible to Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 28

34 compare the findings of the present study with other efforts. From the Results section above, the highest biomass recorded for this project was 1,909 g/m 2 from CD and 973 g/m 2 from Esp, and the mean biomass ranged from g/m 2 (PR) to 6.53 g/m 2 (Nsp). Soltero, et al. (1988) performed a water quality assessment on the euthrophic Eloika Lake, Spokane County, WA, and reported the following mean biomass data: C. demersum 0.73 kg/m 2 (= 730 g/m 2 ), E. canadensis, 0.75 kg/m 2 (= 750 g/m 2 ), P. robbinsii, kg/m 2 (=1,038 g/m 2 ) and P. praelongis 4.42 kg/m 2 (= 4,420 g/m 2 ). In other words, two orders of magnitude higher than the Coeur d Alene Lake results. Chambers and Prepas (1994) characterized the nutrient pool of a Canadian riverbed and indicated that rooted aquatic plants could achieve biomasses greater than 1,000 g/m 2 (although no species were listed) which agrees with the Lower Lakes results. In contrast to the above, Smith and Adams (1986) reported mean shoot biomass of Myriophyllum spicatum to be approximately 225 g/m 2 as a seasonal maximum in Lake Wingra, WI. Similarly, James et al. (2001) studied P. crispus in Half Moon Lake, Eau Claire, WI, and indicated a lakewide June biomass of 31.1 g/m 2 (note: P. crispus is not found in Coeur d Alene Lake but this species senesces in July so the June results approximate seasonal maximum). Filbin and Barko (1985) studied growth and nutrition of macrophytes in Eau Galle Reservoir, WI and showed that the mean seasonal maximum biomass was approximately 300 g/m2 and was comprised primarily of C. demersum and P. pectinatus. Finally, Funk et al (1982) performed a preliminary assessment of the effectiveness of restoration measures at Liberty Lake, Spokane County, WA, and reported maximum standing crop (all species) of between 32 g/m 2 and 257 g/m 2 at various depth in the southern end of the lake. Thus, it appears that the findings of this project in Coeur d Alene Lake are within the range of biomasses seen elsewhere in North America and considerably less than those in euthrophic waters. Nutrient Content As far as nutrient content goes, fewer studies present pertinent comparisons of the phosphorus content of submersed plant species. No comparative data on total nitrogen concentrations was found in the available literature. Again, from the analysis above, the mean TP content ranged from 4,473 μg/g for RA to 2,709 μg/g for PRi. Carignan and Kalff (1982) analyzed a number of species and reported a wide range of TP concentrations in Lake Memphremagog, Quebec. P. richardsonii was the highest at 5,020 μg/g (only one sample) while E. canadensis had a mean of 2,870 μg/g, P. zosteriformis had a mean of 3,160 μg/g and M. spicatum had a mean of 2,500 μg/g. Barko and Smart (1980) reported 3,400 μg/g for M. spicatum in greenhouse-cultured plants. Thus, it appears that the findings of this project in Coeur d Alene Lake are within the range of phosphorus concentration seen elsewhere. Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 29

35 Literature Review of Potential Nutrient Release from Macrophytes Overview The role of macrophytes in the spatial structure and physical-chemical properties of shallow, littoral waters in lakes is widely discussed in the scientific literature (Hutchinson 1975, Wetzel 2001, Jeppesen et al. 1998). Macrophytes are also known to be important in the regulation and cycling of minerals and organic compounds in water bodies (Cooke et al. 2005). Phosphorus is the mineral that is often identified as the limiting nutrient for macrophyte and algae growth (Vollenweider 1970) and since macrophytes may account for a significant fraction of the plant biomass in lakes, it stands to reason that macrophyte nutrition may influence lake-wide phosphorus budgets. It is now recognized that sediment composition exerts an important influence on macrophyte productivity and species composition; however the mechanisms involved are complex (Barko et al. 1991). Lake sediments are reservoirs of nutrients which can be tapped by aquatic macrophytes (Wetzel 2001). Inorganic silt and organic matter are continually accumulated on lake bottoms through the physical settling of suspended material. Both silt and organic matter (plant/algae and animal remains) can bring nutrients to the sediments. Although some of the nutrients incorporated into the sediments may be returned to the overlying water through physical and chemical mechanisms (especially low oxygen and/or high ph conditions in overlying water), most remain there in forms exploitable by aquatic macrophytes. The anoxic, reduced nature of most sediments, typically beneath an oxidized microzone (Barko and James 1997), promotes high solubility of phosphorus and other nutrients (Barko and Smart 1980). Researchers have long debated the relative importance of macrophyte leaves versus roots as primary areas of nutrient uptake (and water versus sediment as primary sources of nutrients). This is an important question related to lake-wide nutrient budgets because macrophytes could be considered as nutrient pumps (i.e. nutrients obtained primarily through the roots and translocated to the leaves during growth) or nutrient sinks (i.e. nutrients absorbed by leaves and stems and thus removed from the water). Early studies seemed to point to roots as primarily attachments to the substrate (Wetzel 2001) but more recent literature supports the view that roots are the main sites of uptake, at least for nitrogen and phosphorus (Barko and Smart 1981, Smith and Adams 1986). The question of water versus sediment as the primary source of phosphorus for macrophytes appears to be put to rest with the findings of Carignan and Kalff (1980) that in both oligotrophic and mildly euthrophic lakes, characterized by relatively high interstitial phosphorus concentrations in the sediments, the sediments constitute the only significant source of phosphorus to rooted plants. Only in rarely encountered hypereutrophic (highly productive) waters is there significant phosphorus uptake from the water. Therefore the relative contribution from water and sediment appears to be a function of their relative phosphorus availability. The pore waters of sediments can be 9 to 600 times richer in phosphorus than the water and thus Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 30

36 rooted macrophytes normally obtain about 85% of the P required for growth directly form the interstitial soil pore water (Carignan and Kalff, 1980). Information on the relative contribution to the nitrogen needed for submersed plant growth is quite limited in comparison to that available for phosphorus. Experiments utilizing the radioisotope 15 N and Myriophyllum spicatum demonstrated that this element can be supplied readily form both water and sediment (Nichols and Keeney, 1976). These experiments also indicated that uptake rates were proportional to the nitrogen concentration in the respective water or sediment and that the subject plant species preferred ammonium over nitrate as the form of nitrogen utilized. Given that ammonium, like phosphorus, is usually in much lower concentrations in the open water, the sediments can be inferred to be the primary source of nitrogen. While the uptake of nutrients by macrophytes has been fairly well determined, the subsequent release of those plant constituents has not. The following discussion presents a summary of available published information on nutrient release from aquatic vegetation, particularly resulting from three primary processes: release from healthy, growing plants, release from materials sloughed from plants during growth and release during plant senescence. The role of algae attached to macrophytes ( epiphytes ) is also mentioned. Macrophyte shoot decay rates depend on several environmental factors, notably temperature, oxygen concentration, nutrient concentrations and chemical composition of the decomposing tissue (Carpenter 1980). These factors were not determined for the Lower Lakes project. The early stages of decay when the plant shoots are still standing in the water column are most relevant to nutrient loading. However, collapsed macrophytes decaying at the sediment surface may lower the redox potential through consumption of available oxygen and thereby enhance diffusion of nutrients from the sediments. This potential, as well, was beyond the scope of this project. Phosphorus Release Phosphorus Release from Growing Plants The scientific literature provides conflicting evidence of excretion of phosphorus by actively growing, rooted aquatic plants. Of those reporting losses to water, few gave estimates of release rates or percentages of cellular nutrients that would or could be released. Differences in study results may be due to experimental conditions or to differences between species. For instance, Carignan and Kalff (1982) found that transfer of phosphorus from nine species of macrophytes (including E. canadensis, M. spicatum, P. richardsonii and P. zosteriformis) to epiphytes was minimal. Similarly, Barko and Smart (1980) tested Hydrilla verticulata, Egeria densa and M. spicatum (all invasive, noxious weeds) and found that cumulative phosphorus released from plant shoots represented less than 10% of the total P mobilized from the sediments. Welsh and Denny (1979, described in Gabrielson et al. 1984) found translocation of phosphorus from roots to shoots and shoots to roots but negligible excretion in Potamogeton. In contrast, Moore et al (1984) used sediments and E. canadensis form Liberty Lake, WA and found that the loss of phosphorus from plants in a laboratory experiment was 25 μg P/gram/day Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 31

37 (assumed to be grams dry weight basis). A previous study by Moore (1981) reported an average of 58 mg P/gram dry weight/day which is substantially higher. These findings are corroborated, in general, by Wallsten (1980, described in Gabrielson et al 1984) that E. canadensis translocated large amounts of phosphorus from roots to foliage with subsequent release to water. For the present study, phosphorus release from growing plants is assumed to be negligible except with E. canadensis, for which a release rate of 25 μg P/gram/day (4.5 mg P/gram/year assuming a 180 day growing season) is used for the nutrient loading calculation. Phosphorus Release from Sloughed Plant Fragments Another factor that can affect potential nutrient release from macrophytes is biomass turnover, or the sloughing off of plant fragments during the growing season. Because this release occurs throughout the growing season, it has a greater potential to influence the growth of algae than the seasonal senescence described below. Defined as annual net production (of a species) divided by the maximum seasonal biomass (of that species), biomass turnover would allow potentially higher levels of nutrient (especially phosphorus) release than senescence. As reported by Carpenter 1980, turnover rates for the larger plants of mesotrophic and euthrophic lakes are generally within the range of 1.0 to 2.6. This means that potentially between 100% and 260% more nutrients could be available for release on an annual basis. Westlake (1975, described in Barko and Smart, 1980) estimated the annual net production of submersed macrophytes to be 1.20 to 1.25 times their seasonal maximum biomass. Smith and Adams (1986) reported that over the entire year 2.8 g P/m 2 was lost from M. spicatum shoots. This was equated to 93% of the plant s total annual phosphorus uptake. However, this magnitude of phosphorus release was indicated to be probably greater than that from other aquatic species based on this plants higher relative productivity and unusually high shoot turnover. Plus milfoil shoot fragments are highly buoyant and can remain suspended in the water until most of the original organic content is lost. Barko et al. (1991) indicate that the effects of aquatic macrophytes on nutrient cycling are most pronounced with the robust submersed species, such as the large species of Myriophyllum, Elodea and Potamogeton because these species have high biomass turnover during the growing season. Unfortunately, no specific release rates were presented by Barko et al., or otherwise found for Elodea or Potamogeton. For the present study, phosphorus release from sloughed plant fragments in Coeur d Alene Lake is assumed to be 1.25 times the seasonal maximum phosphorus content for each species for which nutrient content data was measured for this project. This is the high end of the estimate from Westlake (1975) but towards the lower end of that from Carpenter (1980). This also appears to be conservative compared to the 2.8 g P/m 2 /year presented by Smith and Adams (1986). Phosphorus Release from Plant Senescence In temperate climates many aquatic plants die (senesce) in late summer or fall, some or all of the annual accumulation of biomass is decomposed and nutrients are released (Landers, 1982). By Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 32

38 virtually all accounts this seasonal plant senescence has the greatest potential to provide nutrients from existing plant beds into the open water. However, the process of release, and the influences which mediate this release are extremely complex and experimental results are often conflicting. The stages of aquatic plant decomposition are fairly predictable, according to a review presented by Barko et al (1991). Decay is first initiated by the liberation of soluble materials ( intracellular cytoplasmic compounds ) such as sugars, fatty acids and amino acids. Soluble forms of nutrients are also released, including phosphorus, nitrogen and several cations (sodium, potassium, calcium and magnesium). Leaching and autolysis (splitting open of plant cells) are responsible for rapid, quantitatively significant losses of nitrogen, phosphorus and potassium from plant tissues. Later loss of plant matter is associated with decomposition of more resistant organic materials such as cellulose and lignin. A key influence on the potential release of nutrients from decaying plant matter is the proportion of that matter that is resistant to decomposition; the so-called refractory fraction. Experiments conducted by Jewel (1971) on a number of aquatic species (including E. canadensis and Potamogeton sp. ) resulted in the determination that the refractory fraction varied between 11% and 50% and averaged 24% of the initial organic matter content. Horne and Goldman (1994) indicate 30% as the refractory fraction of phosphorus in the phytoplankton (algae). Jewell further indicates that the quantities of nitrogen and phosphorus regenerated (released) during decay can be predicted within 25% by using the measured refractory fraction and the initial nutrient concentration. From his studies the species-specific refractory fraction for E. canadensis was 19.2% and 27.4% (unaltered samples) and that for Potamogeton species was 19.3% (chopped samples). In these same trials Jewell determined that 78.6% of the phosphorus was released from Elodea and 100% of the phosphorus was released from Potamogeton sp. (thus the simple refractory fraction approach for these two species appeared to underestimate the actual release slightly). Using data presented in Nichols and Keeney (1973), Landers (1982) calculated that 27% of the phosphorus and 53% of the nitrogen should remain in decayed M. spicatum as a refractory fraction. From this Landers concluded that approximately 70% of the phosphorus and 50% of the nitrogen present in the plant tissue before senescence would be released to the surrounding water. However, Nichols and Keeney indicate in their 1973 publication that, from their experiments, the presence of sediments in the experimental containers caused much less phosphorus to be released due to absorption of dissolved, inorganic phosphorus compounds. In another laboratory study, this time using Ceratophyllum demersum, Best et al (1990) monitored plant decomposition and found that, under aerobic conditions only between 3.1% and 5.7% of the initial total phosphorus content was found in the water; the majority was in the sediments or remaining leaf litter. Another factor, referred to by several researchers is the time between the onset of senescence and when the plant collapses onto the sediments, at which time nutrient release from the plants to the open water essentially ends (Nichols and Keeney, 1976; Carpenter, 1980). This is undoubtedly affected by many factors, such as temperature and plant species. In Coeur d Alene Lake the Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 33

39 influence of drawdown (which could potentially shorten the time to collapse by removing the water column) is essentially an unquantifiable factor. For the present study, phosphorus release from seasonal submersed plant senescence in Coeur d Alene Lake is assumed to be 50% of the measured phosphorus content. Phosphorus Release from Epiphyton A potentially significant influence on phosphorus or nitrogen that might be released by macrophytes (during active growth sloughing or senescence) is that of epiphyton (attached algae and other micro-organisms). Calignan and Kalff (1982) studied the contribution of nine species of macrophyte to the phosphorus nutrition of their epiphyton and found that the epiphytes derived less than 10% of their phosphorus from the supporting macrophytes. Thus epiphytes obtain most of their phosphorus from the open water. This study found that very little phosphorus would be transferred directly from the macrophytes to the surrounding waters; instead it would be released indirectly via epiphyte metabolism. This infers another, perhaps relatively minor, error in the attempt to perform a simple calculation of nutrient release from macrophytes. For the present study, phosphorus release from epiphyton is considered an un-quantified factor. Nitrogen Release The literature search performed for this project obtained little usable information on the release of nitrogen from macrophytes, especially during active growth. Nichols and Keeney (1973) found that nitrogen tends to be retained or accumulated on particulate macrophyte matter and that release of dissolved nitrogen forms (such as ammonium and nitrate) is minimal. Nichols and Keeney also indicated that nitrogen is required in the decay process, assumably by the organisms that contribute to this decay. The idea that a refractory fraction of plant biomass will limit nutrient release in general is likely pertinent to this analysis; however, it appears from the analysis above regarding phosphorus release that there are other more restrictive limitations to this. Jewel (1971) reported that no release of nitrogen resulted from the senescence or sloughing of Potamogeton species while that from E. canadensis was between 42.8 and 64.5%. Best et al. (1990) followed plant decomposition of C. demersum and found that between 3.0 and 6.8% of the plant s nitrogen was released under aerobic conditions. Given the level of uncertainty surrounding the release of nitrogen from aquatic vegetation, nitrogen release from seasonal submersed plant senescence in Coeur d Alene Lake is assumed to be 10% of the measured total nitrogen content for all species. Following the discussion above regarding phosphorus release from sloughed plant materials, a rate of 1.0 times the seasonal maximum nitrogen content is assumed for all species. Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 34

40 Estimate of Nutrient Loading from Aquatic Vegetation Overview Effects of submersed aquatic macrophytes on nutrient cycling in lakes are most pronounced in shallow areas that support extensive stands of robust submersed species, such as the large species of Myriophyllum, Potamogeton and Elodea (Carpenter, 1980). These species have high biomass turnover during the growing season and therefore recycle nutrients when water temperatures are high and potential effects on plankton production are maximal (Carpenter, 1983). This being said, the estimation of nutrient release from a diverse plant assemblage in a lake the size of Coeur d Alene, or even the Lower Lakes study area, is a complex process and involves many assumptions to expand the limited available data. There are numerous sources of uncertainty, and thus error, in this type of estimation that should be noted. The sampling error is relatively high not because of errors in the sample collection but because of the inherent patchiness of aquatic vegetation in lakes. This is compounded by the expanse of this lake and the variation in bottom slope and substrate, ambient water clarity, aspect etc. There is also analytical error which potentially affects the results. The lab data reports indicate QC limits of 20% on biomass and nutrient analyses but review of the QC data indicates that this actual error is minimal (less than 5%). There is an obvious error in using averaged biomass and nutrient concentrations and extrapolating these across wide areas. The use of averages was deemed necessary given the somewhat limited time available to complete the data analysis and project report. However, perhaps the greatest uncertainty in this analysis is in the nutrient release criteria. As was indicated above, it was hoped to obtain species-specific phosphorus and nitrogen release factors to be used in the loading calculations. However, the only species-specific value that was found was for the release of phosphorus during the growth of Elodea, and that was provided by only one study. Further literature search and review might help refine the estimate produced but it appears that the needed information is simply not available. In spite of the expressed error factors, it is believed that the estimates produced for this project are sound and reasonable. Aquatic Plant Growth Regions Areas of aquatic plant growth were grouped for this analysis into areas (regions) which were reasonably homogeneous (in terms of depth, bottom type and species present) and represented by the sampled transects. A description of location and area of these regions is presented in Table 10 and the regions are shown in Figure 9. The surface area of each region was developed from Tribal GIS Program depth contour mapping data. Basically, a series of polygons were manually

41 Table 10. Aquatic plant growth regions established for the Lower Lakes Aquatic Vegetation Survey project. Lower Lakes Average Aquatic Vegetation Representative Maximum Depth Area Region Transects of plant growth (ft) (acres / m 2 ) Benewah Lake B1, B2, B3, B4 and B / 886,689 East Chatcolet Lake C6, C9, C10 and C11 No limit 413 / 1,672,913 North Chatcolet Lake C1, C7 and C / 1,156,349 SW Chatcolet Lake C2, C3, C4, C5 and C / 604,547 Round Lake R1 through R7 No limit 891 / 3,606,420 Figure 9. Location of aquatic plant growth regions used for nutrient release estimation for the Lower Lakes Aquatic Vegetation Survey project. Project Completion Report for the Lower Lakes Aquatic Vegetation Survey Project 36